In this paper an object recognition and tracking approach for the mobile, marker-less and PDA-based augmented reality system AR-PDA is described. For object recognition and localization 2D features are extracted from images and compared with a priori known 3D models. The system consists of a 2D graph matching, 3D hypothesis generation and validation and an additional texture based validation step.

The paper presents "vertex tracing", a new approach to adaptive progressive ray tracing. The software achieves real-time rendering performance of physically accurate reflection effects on specified scene objects, even with complex scenes in virtual reality applications. The approach is based on hybrid rendering, combining an OpenGL-generated scene with correct reflection characteristics of selected scene objects. The real-time performance of the vertex tracer is achieved with progressive adaptive refinement of the geometry in object space, and with parallelization of the algorithm. Both shared memory and distributed memory architectures have been investigated. Mesh-based load balancing yields a uniform distribution of the computing load, also in a heterogeneous network with resources of widely varying performance. The performance of the overall system is demonstrated with a truck interior in a virtual reality simulator.